Cross-selling strategies

Oct 01, 2008 9:30 PM  By

Whether the product is fast food, weddings, cars or peanut butter, cross-selling can increase profits and lift natural search traffic. Cross-selling, a sales technique used to inform a customer of an item related to the product in focus, is best defined by the printer purchase: Big box stores typically cross-sell printer customers, reminding them that they should also pick up paper, ink and a network or USB cable.

How does cross-selling apply to search engine optimization (SEO)? The answer lies in facilitating the discovery of complementary, new and bundled products for both customers and search robots.

Keyword themes and product bundles


Bundling can both cross-sell multiple products and, in some cases, even create a semantic relationship in search engines to other products. A common cross-selling technique is to bundle several related products together and then drop the price below what the total would be if purchased separately. The customer sees enhanced value and gets introduced to other products he or she may have otherwise not purchased, all while raising the average order.

Performing a simple tilde search will illustrate this point. If you want to search not only for your search term but also for its synonyms (as Google sees it), place the tilde sign (“~”) immediately in front of your search term.

For example, a search for [~office] in Google will produce bolded results that include words like “secretary,” “Outlook” and “Excel.”

Achieving a semantic connection amongst product lines is possible. A tilde search on the [~Elantra] will also return the Hyundai Sonata and Tiburon. BMW is in the fortunate situation of having a strong, unbranded semantic relationship with the Google query [~car].

Using the bundling cross-selling technique and interlinking these different product pages creates value in two distinct ways. First, bundling raises the average order value (AOV). Second, bundling creates a semantic relationship amongst a product family in search engines, Google in particular.

So when someone searches for Product A, Google will automatically associate Products B and C with it. This can result in Google actually cross-selling the product family before the user even gets to the site. When a customer searches about a single item, Google sometimes includes additional semantically related search terms. The query [office] includes the related search terms [office depot], [office max] and [office pranks]

You can test your product line to see what products are your semantic cousins either by using the tilde search or by visiting Google Sets at http://labs.google.com/sets. Entering [accord, insight] at Google Sets will result in Google showing the entire Honda vehicle lineup.

Natural search benefits


Why is cross-selling complementary products an effective SEO strategy? Because of the text links that cross-selling generates. How you interlink your pages determines to a large extent how much link authority (PageRank) each of your pages inherits. A page that is higher up in the “site tree” (think: org chart) will command more respect from the search engines.

Also, the number of internal links factors in to the ranking equation. So a page with only one internal category-level link isn’t likely to fare as well as one that receives many such links. Another ranking factor is the underlined text of the links pointing to the page — and that’s true whether it’s an internal link or an inbound link. This is called “anchor text.” Linking with relevant anchor text will create semantic relationships between your products — a good thing for SEO. Interlinking product pages can forge a powerful semantic bundle.

Think of it this way: When a product or service performs better with a complementary product, simply add a link to that cross-sell using relevant keywords to help it do better in the search engines — in the form of indexation, ranking and traffic benefits. Use your more popular pages to introduce product bundles to both raise the AOV and create higher perceived value.

Split-testing bundles


Use split testing to see which bundles of complementary products are most profitable by comparing the margin from the bundle multiplied by the conversion rate of the bundle. While bundling Product A with high-end B is more profitable, it may have a lower conversion rate than a bundle of Product A and lower-end C. Use a split test to see which combination of A, B and C produces the maximum profit by analyzing margin and conversion rate.

The resulting data can help you find the optimum cross-selling opportunity for incorporating these terms into your natural search strategy and for creating long tail keyword bundles. If [peanut butter] is highly competitive, try optimizing for [peanut butter jelly]; it might be more targeted and have less competition. These bundled terms could also be low cost/high conversion PPC phrase match candidates. Google defines phrase match as more targeted than broad match, but more flexible than exact match.

What if we don’t have enough data to know which products to bundle?


Well, there is a repository of data online, and you have probably already paid it a visit. Amazon.com has graciously shared lots of its compiled data stemming from many years of empirical research. For some time Amazon actually published the percent of people who viewed the current product and what they subsequently purchased:

Of the people who viewed Product A

  • 59% buy Product B
  • 32% buy Product C
  • 14% buy Product D

While Amazon no longer provides the percentages, it does show you a multitude of bundle combinations and complementary products. When you are considering what product types to bundle, Amazon should be a first stop. It is probably safe to assume that Amazon is bundling the products with the highest chance of conversion success.

Using a Web analytics package to measure a searcher’s intent, click through and click path can help locate effective cross-sells as well.

Recommendation engines, cross-selling and SEO


Recommendation engines help consumers make an informed purchasing decision, and serve online merchants in three distinct ways.

  • First, they generate unique content. Amazon has effectively “crowd sourced” a wealth of unique Web content that is highly topically relevant and effectively targets many long tail search queries around any given product.

  • Second, these reviews can serve as opportunities to incorporate keyword-rich text links to cross-sell other items within the site. Not only will this provide a path for consumers to click through, it will also provide a rich semantic trail for search engine spiders to follow.

    The more keyword-containing links a page receives, the stronger the natural search boost in search engines for that particular keyword. If 20 pages within one site point to a page about “the effects of peanut butter on dogs,” then search engines will subsequently deduce that, within this site, this is the authority page on that subject matter.

    Interlinking can pay large dividends by thematically distinguishing pages within a large e-commerce site. Every additional signal sent to search engines about what the leading internal page is for a given subject matter helps.

  • Third, by successfully testing and tracking conversion success based on recommendations, merchants can show customers a high conversion item earlier in the session — before they leave the site.

Recommendation engines, collaborative filtering and behavioral targeting technology can be of immense help in providing recommendations for related content and/or products. This can be based on traffic patterns, click-through rates, past purchases and other indicators of consumer behavior.

Tim Gill is a natural search analyst and Stephan Spencer is founder/president of Netconcepts (www.netconcepts.com), a natural search marketing firm based in Madison, WI.